def testSubsituteNonExistChar(self):
tokens = ['AAAAA', 'KKKKK']
aug = OcrAug()
for t in tokens:
augmented_text = aug.augment(t)
self.assertEqual(t, augmented_text)
self.assertTrue(len(tokens) > 0)
def test_ocr_single_word(self):
texts = ['Zoology', 'roku123456']
aug = OcrAug()
for text in texts:
augmented_text = aug.augment(text)
self.assertNotEqual(text, augmented_text)
self.assertTrue(len(texts) > 0)
def testSubsituteExistChar(self):
tokens = ['Zoology', 'roku123456']
aug = OcrAug()
for t in tokens:
augmented_text = aug.augment(t)
self.assertNotEqual(t, augmented_text)
self.assertTrue(len(tokens) > 0)
def test_ocr_single_word_nonexist_char(self):
texts = ['AAAAA', 'KKKKK']
aug = OcrAug()
for text in texts:
augmented_text = aug.augment(text)
self.assertEqual(text, augmented_text)
self.assertTrue(len(texts) > 0)
def test_ocr_model_from_json(self):
sample_path = os.path.abspath(
os.path.join(os.path.dirname(__file__), '..', '..', 'res',
'common', 'sample.json'))
aug = OcrAug(dict_of_path=sample_path)
augmented_text = aug.augment('0000000')
self.assertIn('3', augmented_text)
with self.assertRaises(Exception) as error:
sample_path = os.path.abspath(
os.path.join(os.path.dirname(__file__), '..', '..', 'res',
'common', 'non_exist.json'))
aug = OcrAug(dict_of_path=sample_path)
self.assertIn('The dict_of_path does not exist', str(error.exception))
def test_ocr_multi_words(self):
texts = ['The quick brown fox jumps over the lazy dog']
aug = OcrAug()
for text in texts:
# Since non-exist mapping word may be drawn, try several times
is_augmented = False
for _ in range(10):
augmented_text = aug.augment(text)
is_equal = text == augmented_text
if not is_equal:
is_augmented = True
break
self.assertTrue(is_augmented)
self.assertTrue(len(texts) > 0)
def test_empty_input(self):
# Empty input
tokens = ['']
aug = OcrAug()
for t in tokens:
augmented_text = aug.augment(t)
self.assertEqual(augmented_text, '')
self.assertEqual(len(tokens[0]), 0)
self.assertTrue(len(tokens) > 0)
tokens = [None]
aug = OcrAug()
for t in tokens:
augmented_text = aug.augment(t)
self.assertEqual(augmented_text, None)
self.assertEqual(len(tokens), 1)
def test_ocr_model_from_dict(self):
mapping = {'0': ['2']}
aug = OcrAug(dict_of_path=mapping)
augmented_text = aug.augment('0000000')
self.assertIn('2', augmented_text)
def test_ocr_empty(self):
texts = ['', None]
aug = OcrAug()
for text in texts:
augmented_text = aug.augment(text)
self.assertEqual(text, augmented_text)
